Université de Montréal
Effets protecteurs du Celsior sur la fonction endothéliale d’artères
coronaires porcines dans un modèle de préservation myocardique
et de greffe cardiaque hétérotopique
Par
Eric Durnont
Département de Chirurgie
Faculté de Médecine
Mémoire présenté à la Faculté des Études supérieures en vue de
Fobtention du grade dc Maîtrise en Sciences Biomédicales
Juin, 2003
/
Grade conféré / àcompterdu •a\© Eric Dumont, 2003
1004 MR. I\
u
Lf
056
0
ooLi vl 13flo
o
de Montréal
Direction des bibliothèques AVIS
L’auteur a autorisé l’Université de Montréal à reproduire et diffuser, en totalité ou en partie, par quelque moyen que ce soit et sur quelque support que ce soit, et exclusivement à des fins non lucratives d’enseignement et de recherche, des copies de ce mémoire ou de cette thèse.
L’auteur et les coauteurs le cas échéant conservent la propriété du droit d’auteur et des droits moraux qui protègent ce document. Ni la thèse ou le mémoire, ni des extraits substantiels de ce document, ne doivent être imprimés ou autrement reproduits sans l’autorisation de l’auteur.
Afin de se conformer à la Loi canadienne sur la protection des renseignements personnels, quelques formulaires secondaires, coordonnées ou signatures intégrées au texte ont pu être enlevés de ce document. Bien que cela ait pu affecter la pagination, il n’y a aucun contenu manquant. NOTICE
The author of this thesis or dissertation has granted a nonexclusive license allowing Université de Montréal to reproduce and publish the document, in part or in whole, and in any format, solely for noncommercial educational and research purposes.
The author and co-authors if applicable retain copyright ownership and moral rights in this document. Neither the whole thesis or dissertation, nor substantial extracts from it, may be printed or otherwise reproduced without the author’s permission.
In compliance with the Canadian Privacy Act some supporting forms, contact information or signatures may have been removed from the document. While this may affect the document page count, it does not represent any loss of content from the document.
Université de Montréal
Faculté des études supérieures
Ce mémoire intitulé
Effets protecteurs du Celsior sur la fonction endothéliale d’artères
coronaires porcines dans un modèle de préservation myocardique
et de greffe cardiaque hétérotopique
Présenté par:
Eric Durnont
A été évalué par un jury composé des personnes suivantes:
Michel White
Président-rapporteurMichel Pellerin
Directeur de rechercheLouis P. Perrault
Co-directeur de rechercheDenis Bouchard
Membre dujuryLa survie des patients après transplantation cardiaque est limitée par l’apparition d’une forme d’athérosclérose accélérée au niveau des artères coronaires menant à des infarctus silencieux et une perte progressive de la fonction du greffon. L’apparition précoce d’une dysfonction endothéliale au niveau des artères coronaires prédit le développement de la maladie du greffon 1 an après la transplantation. Les facteurs impliqués dans l’apparition de cette dysfonction endothéliale incluent le rejet cellulaire et humoral, la dyslipidérnie, l’infection au cytomégalovirus, l’exposition des artères coronaires aux solutions cardioplégiques, l’ischémie froide, et le type de solution de préservation utilisé lors du prélèvement cardiaque.
Le premier article rapporte l’effet de l’utilisation du Celsior, une nouvelle solution de préservation spécifiquement conçue pour la greffe cardiaque, sur la fonction endothéliale d’artères coronaires porcines dans un modèle de préservation cardiaque comparé à deux stratégies courantes de cardioplégie et de préservation utilisées en greffe cardiaque. Les résultats obtenus démontrent que l’utilisation du Celsior pour l’induction de la cardioplégie et la préservation myocardique a un effet protecteur à court terme sur les relaxations endothélium-dépendantes médiées par les protéines G comparé aux autres stratégies. Ces résultats pourraient être associés avec
des meilleurs résultats à court et moyen terme en greffe cardiaque avec l’utilisation du Celsior lors du prélèvement.
Le deuxième article rapporte les effets protecteurs du Celsior sur la fonction endothéliale des artères coronariennes épicardiques porcine à 30 jours post-transplantation cardiaque hétérotopique chez le porc. L’utilisation du Celsior pour l’induction de la cardioplégie et la préservation est comparé avec deux stratégies de cardioplégie et préservation utilisées en greffe cardiaque. Les résultats obtenus démontrent une meilleure réactivité endothéliale des artères coronaires épicardiques préservées au Celsior 1 mois après la transplantation. L’utilisation du Celsior pour l’induction de la cardioplégie et la préservation pourrait potentiellement réduire l’incidence d’athérosclérose accélérée du greffon comparé aux autres solutions en préservant les effets protecteurs de l’endothélium sur la paroi vasculaire.
Mots clés : endothélium, athérosclérose précoce, cardioplégie, préservation, greffe cardiaque.
Suiwival afier heart transplantation is lirnited by the appearance of an accelerated form of atherosclerosis in the coronary arteries that leads to
suent infarcts and progressive lost of grafi function. Early appearance of coronary artery endothelial dysfunction is predictive of deveÏopment of coronary allografi vasculopathy one year afier transplantation. Factors which have been incriminated as the cause of this dysfunction include cellular and humoral rejection, hypercholesterolemia, CMV infection, exposure of coronal-y arteries to cardioplegic solutions, cold ischemia, preservation solutions, and reperfusion afier implantation.
The first article evaluates the effects of Celsior (an antioxidant solution specifically designed for cardiac preservation) in a model of heart preservation (4 hours to mimic clinical conditions) cornpared to two commonly used cardioplegic and preservation strategies on coronary artery
endothelial function in a porcine model. The resuits demonstrate that Celsior cardiopÏegia and storage allowed for functional recovery of endothelium dependent relaxations to serotonin cornpared to saline preservation. The observed effect may be associated with an improvement in both short- ad long-term outcome in heart transplantation.
The second article examined the potential protective effects of Celsior, a new preservation solution specifically designed for heart transplantation, on coronary endothelial reactivity at 1 month post transplantation cornpared with crystalloid and blood based cardioplegic solutions. The results dernonstrated that the use ofcrystalloid solution causes a decrease in endothelium-dependent relaxations to agonists as compared to blood and Celsior solutions. The use of Celsior has a protective effect on endothelial function at 1 rnonth and potentially reduces the incidence of coronary allografi vasculopathy cornpared to other solutions by preserving the protective effects ofthe endothelium on the vascular wall.
Key words: Endothelium, accelerated atherosclerosis, cardioplegia, preservation, heart transplantation.
TABLE DES MATIÈRES Page PAGE TITRE 1 RÉSUMÉ (FRANÇAIS) 2 RÉSUMÉ (ANGLAIS) 4 CONTENU
Liste des figures 7
Liste des abbreviations 8
REMERCIEMENTS 9
INTRODUCTION 10
ARTICLE 1 29
Improved preservation of coronary endothelial function with Celsior cornpared with blood and crystalloid solutions in heart transplantation
ARTICLE 2 61
Protective effects of Celsior on coronary endothelial function one rnonth after heterotopic porcine heart transplantation
CONCLUSION 98
8
LIST 0F FIGURES AND TABLES
INTRODUCTION
figure 1 Endothelium-derived relaxing factors
figure 2 Relationship between EDRF and rise in intracellular calcium concentration
figure 3 Neurohormonal mediators
figure 4 Signal transduction pathway in a normal endothelial ceil Figure 5 Interaction between platelet, thrombin, and the
endothelium
Table 1 Composition and characteristics of Celsior solution
ARTICLE 1
Table 1 Composition and characteristics of Celsior solution Table 2 Experirnental groups
Figure 1 Amplitude of contractions to potassium chloride
f igure 2 Amplitude of contractions to PGf2Œ
f igure 3 Concentration-response curve to serotonin (5-HT) and bradykinin (BK) afler cardioplegia alone
Figure 4 Concentration-response curve to serotonin (5-HT) Figure 5 Concentration-response curve to bradykinin (BK)
ARTICLE 2
Table 1 Experimental groups
f igure 1 Cumulative concentration-response curves to serotonin (5-HT)
Figure 2 Cumulative concentration-response curves to bradykinin (BK)
Figure 3 Cumulative concentration-response curves to calcium ionophore (A23 1 87)
ABBREVIATIONS
5HT Serotonin
Ach Acetylcholine
ANOVA Analysis of variance ATP Adenosine triphosphate
BK Bradykinin
Ca++ Calcium
A23 187 Calcium ionophore
cAMP Cyclic adenosine 3,5 -monophosphate cGMP Cyclic guanosine 3,5-monophosphate EDCF Endothelium derived contracting factor
EDHF Endothelium derived hyperpolarizing factor EDRF Endothelium derived relaxing factor
eNOS Endothelial nitric oxide synthase ICAM Intracellular adhesion molecule iNOS Inducible nitric oxide synthase 1P3 Inositol triphosphate
ISHLT International Society for Heart and Lung Transplantation
IVUS Intravascular ultrasound
L-NMIVIA N-monomethyl-L-arginine
MHC Major histocompatibility complex
NADPH Nicotinarnide adenine dinucleotide phosphate nNO$ Neuronal nitric oxide synthase
NO Nitric oxide
NOS Nitric oxide synthase
PDGF Platelet derived growth factor PGF2ΠProstaglandin F2 alpha
PGI2 Prostaglandin 12 (Prostacyclin)
Gi Protein Gi-coupled protein (GIPase binding protein) Gq Protein Gq-coupled protein
SEM Standard error of the mean SMC Smooth muscle cells
SNP Sodium nitroprussiate TXA2 Thromboxane A2
UW University of Wisconsin solution
‘o
Remerciements
J’aimerais remercier mes directeurs de recherche, Dr Louis P. Perrault et Dr. Michel Pellerin pour m’avoir fait confiance dans ce projet, pour leurs précieux conseils, et pour leur dévouement à me guider et me corriger.
Je ne peux oublier l’aide extraordinaire que m’a apporté Nathalie Desjardins pour les manipulations et greffes, les chambres d’organes et la tabulation des statistiques. Sans elle, mon séjour au laboratoire aurait été
différent et cet ouvrage n’aurait jamais vu le jour.
Merci aussi à Olivier et Josie pour leur aide et leur présence dans le laboratoire. Je vous garde une place dans ma mémoire.
Finalement,
j
‘aimerais remercier mes parents qui m’ont toujours supporté durant ces longues études et Leah pour sa patience, son amour et sa générosité. Hang in there, Babe!Introduction
The Normal Endothehum
The normal endothelium contributes to vasomotor tone and to the maintenance of a non-thrornbogenic vascular surface, acts as a selective barrier for the transport of macrornolecules and solutes needed for cellular metabolisrn, contributes to the proliferation of the underlying smooth muscle, and regulates the adhesion and transmigration of neutrophils, monocytes and lymphocytes. These endothelial properties are due to its capacity to sense hormonal and hemodynamic stimuli by three basic rnechanisms: the secretion of endothelium-dependent factors, the expression on its surface of celi adhesion molecules and signal transduction coupling molecules, and morphological changes.
11e role of the vascular endothelium was first recognized by Furchgott and Zawadski in 1980 (1) following the discovery of the endothelium-dependency ofthe dilator response to acetylcholine. Since then, three pathways have been described that mediate endothelium-dependent relaxations. Endothelium-dependent vasodilatation can be mediated by the
12
release of nitric oxide (NO), prostacyclin, or endothelium-dependent hyperpolarizing factor (EDHF) (Figure 1) (1-6).
1) Nitric Oxide
Nitric oxide is a labile molecule and diffuses readily through plasma membranes. It is formed from L-arginine via the constitutive form of nitric oxide synthase (7,8) (NO synthase III). The activation of this enzyme depends on the intracellular concentration of calcium ion in the endothelial ceil, on calmodulin, and requires nicotinamide adenine dinucleotide phosphate (NADPH) for optimal activity (9). NO synthase can be competitively inhibited by analogs of L-arginine N-monomethyl-L-arginine (L-NMIvIA) and N-nitro-L-arginine (Figure 1) (9). There are three types of nitric oxide synthase enzymes responsible for the production of NO. Nitric oxide synthase type I (nNOS) exists in the neuronal cells of the brain in peripheral nonadrenergic noncholinergic neurons (10). It produces NO for the activation of the signal transduction pathways in neurotransmission. Inducible NOS II (iNOS) is produced by macrophages in response to endotoxins, rnicro-organisms, or cytokine secretion. INOS has shown cytotoxic effects on invading bacteria and tumor celis, is capable of producing much greater quantities of NO than the other isoforms (11), and is responsible for the vasodilatatory effect seen in sepsis and septic shock.
Endotheliai ceils generate NOS III (eNOS) which produces NO one of several endothelium-derived relaxing factor (EDRF). Most vasoactive stimuli such as bradykinin and shear stress mediate their activities through the production of NO derived from the endothelium. Ccli membrane-bound G-proteins transduce signais intracellularly and play a pivotai role in celi bioiogy (12-14). These intraceilular signais cause an increase in inositol triphosphate and diacylglycerol which in tum alter calcium leveis within the ceil and activate Ca++ dependent proteins such as eNOS. NO diffuses into smooth muscle ceils within 200 jim and causes relaxation via activation of guanylate cyclase in vascular srnooth muscle (15) (Figure 2). Since NO has a short half-life of 6 seconds, its effects are stimulus and dose-dependent. Once entered into the smooth muscle cdl (SMC), NO binds to soluble guanylate cyclase which stimulates the conversion of GTP to guanosine 3’,5’-monophosphate (cGMP). cGMP is associated with inhibition of $MC contractility by shifling the intracellular calcium (Ca++) to a lower concentration. Since an increased concentration of Ca++ is required for the dephosphorylation of the rnyosin light chain kinase and thus smooth muscle contraction, rises in the cGIVW concentration induce vasorelaxation (16-1 8). The production of NO significantly contributes to endothelium-dependent relaxations in coronary, systemic, mesenteric, and pulmonary beds. Tts
14
contribution to vasomotor tone is suggested by the fact that inhibitors of NO cause a generaiized vasoconstriction and a rise in arterial bÏood pressure in animais and humans (19,20). Endothelial ceils also secrete NO in the lumen of biood vesseis which is imrnediately neutralized in normai situations by the presence of oxyhemoglobin. Thus NO acts as an interface between the blood stream and the endothelium and in this way inhibits platelet and leukocyte adhesion. Physiologically, platelets norrnally circulate in the peripheral bloodstream in an inactivated state due to the liberation of NO and prostacyclin by the endothelium. Both mediators increase cGMP or cAIVW in piatelets and thereby prevent platelet adhesion and aggregation (21-24). At sites where plateiets are activated, they release active mediators which interact with endothelial receptors (Figure 5) and cause endothelium dependent relaxations mediated by NO. The presence of endotheiium inhibits to a large degree the vasoconstriction induced by thrornboxane A2 and serotonin derived from platelets during aggregation. The piateiet-derived mediator primarily responsible for the stimulation of NO formation is ADP and to a lesser extent serotonin, which act respectively on Py-purinergic and 5-HT1d receptors (25,26). Serotonin aiso binds to 5HT2 receptors on smooth muscle ceils which cause vasoconstriction when stimuiated. The action of platelet products and thrombin is crucial for the protective role of the
endothelium against inappropriate coagulation. Not onÏy does the release of NO induce a direct relaxation of srnooth muscle thereby inducing vasodilatation and preventing platelet microaggregation, but the endothelial barrier prevents the vasoconstricting substances (thrornboxane A2 and serotonin) from acting on the vascular srnooth muscle. When the endothelial barrier is darnaged, platelet aggregation goes unopposed with the liberation of thromboxane A2 and serotonin and the platelet activator effect of thrombin (25).
Numerous neurohormonal stimuli can cause the release of nitric oxide from endothelial ceils (Figure 3). Circulating hormones such as catecholarnines or vasopressin, prostanoids such as histamine or bradykinin, mediators derived from pÏatelets (serotonin, ADP) or forrned during coagulation (thrornbin) can ah bind to specific receptors on the surface of endothelial cells and cause the activation of eNO$ via distinct signal transduction pathways (Figure 4). For example, Œ-adrenergic, serotonin, and thrornbin receptors are coupled to a Gi-protein pathway which is sensitive to pertussis toxin whereas ADP and bradykinin (which also mediates release EDHF) mediate production of NO via activation of Gq-proteins (27,2$).
Activation of both these protein pathways by the binding of agonists to their cell surface receptors leads to an intracellular increase in inositol
16
triphosphate and diacylglycerol, an increase in intracellular calcium and activation of eNOS. The Gi-protein mediated pathway is one of the first pathways affected in many cardiovascular diseases such as acute and chronic cardiac rejection (29,30) and atherosclerosis (31,32), hypercholesteroïemia, and ischernia-reperfusion injury and may be an early marker ofpathological activation ofthe endothelium in vascular diseases. The Gq-protein rnediated pathway is more robust and is usually affected later in the disease process. 2) Prostacyclin (PGI2)
Prostacyclin is predominantly forrned by the endothelium, although vascular smooth muscle ceils have a lesser capacity to produce prostanoids. It is a major byproduct of cyclooxygenase and induces SMC relaxation in response to hypoxia and shear forces. Prostacyclin causes vasodilatation by increasing cyclic AMP in vascular smooth muscle cells. In rnost vessels, the contribution ofprostacyclin to endothelium-dependent relaxation is minimal but it acts synergistically with NO to prevent platelet aggregation (33,34). 3) Endothelium-derived Hyperpolarizing Factor
Recent studies have demonstrated that acetylcholine and other endothelium-dependent vasorelaxants cause an endothelium-dependent hyperpolarization from a diffusible factor derived from the endothelium (endothelium derived hyperpolarizing factor) distinct from NO and
prostacyclin (35-39). Indeed, inhibitors of nitric oxide formation or cyclooxygenase such as L-NMMA or indornethacin only partially inhibit endothelium-dependent relaxations to bradykinin. The chemical nature of EDHF remains to be determined but it causes hyperpolarization of vascular smooth muscle ceils via ATP-dependent potassium channels (40-42). The exact type of potassium channel is unknown but is probably calcium dependent. The contribution ofEDHF to endothelium-dependent relaxation varies according to the size of the vesse! but is greatest in resistance arteries
(38,43).
4) Endothe]ium-derived Contracting Factors
The endothelium can a!so produce contracting factors derived from the cyclooxygenase pathway, narnely prostaglandin H2, arachidonic acid, and thromboxane A2 (44), as wel! as endothelin-1(44-49). Endothelin-1 is more likely to play a role in the long term regulation of vasomotor tone rather than the spontaneous vasoconstriction of the vascular smooth muscle. Endothe!in production is stimulated by chemical stimuli such as hypoxia, and mechanica! forces, as well as agonists such as thrombin, vasopressin, and angiotensin II.
1$
Coronary Allograft Vasculopatliy and Endothelial Dysfunction
Heart transplantation remains the only efficacious treatment for terminal heart failure but has an associated 5-year survival of 75% owing to the developrnent of graft coronary vasculopathy, which leads to suent infarcts, grafi failure, and sudden death (50-52). It is estirnated that over 50% of patients have significant stenotic disease at 3 years afier transplantation using intravascular ultrasound for evaluation and 50% of these will develop graft failure (53). Most ofthese individuals do not present symptoms of the disease because of cardiac denervation occurring during harvesting. The first clinical manifestations are often arrythmias, heart failure, or sudden death. Grafi coronary vasculopathy is manifested by a unique and unusually accelerated form of coronary disease affecting both the smaller intramyocardial vessels and the larger epicardial arteries (54). Cardiac allografi vasculopathy affects only allograft vessels, sparing ail native vesseis. Intimai hyperplasia is visible in up to 75% of patients at 1 year using IVUS (55).
The pathogenesis and morphological pattems of coronary allograft vasculopathy are yet to be fully elucidated but the histological changes have been described. Pathological examination of coronary arteries from human cardiac allografts have shown that compared with conventional
atherosclerotic plaques, the disease is diffuse rather than segmental,
concentric in nature rather than eccentric, deveÏops within months to years,
and also occurs in the pediatric population afier transplantation (56,57-61).
The initial lesion is characterized by an early intimai proliferation which
progresses with time to plaques incorporating lipid deposits and late
calcification of coronary vessels (54). These histopathological changes are a
function of tirne; earÏy afier transplantation a vasculitis predominates
characterized by cellular infiltrates consisting of rnodified smooth muscle
ceils, T lymphocytes, and rnonocytes/macrophages which progress to focal
atherosclerotic plaques (59,62). The intimai hyperplasia is diffuse which
explains the obliteration of the small distal resistance vessels before the
larger epicardial coronary arteries (63).
Coronary endothelial dysfunction has been shown to be a precursor to
the developrnent of intimai hyperpiasia and atherosclerosis in many vascular
diseases such as hypercholesterolemia and diabetes. It also precedes the
development of coronary allograft vasculopathy as is shown by an impaired
endothelium-dependent relaxation to different stimuli at different time points
afier transplantation: the vasomotor response to acetylcholine (64,65) and
the cold pressor test (66) may be abnormal a few weeks to rnonths afier
20
for several months to years (67). These abnormalities are predictive of the developrnent of graft vasculopathy and may predict an adverse clinical outcome (50,68-69). Endothelial ceil damage also leads to enhanced release of endotheÏin, a potent vasoconstrictor peptide, with subsequent increase in coronary vascular resistance and diminished myocardial blood flow (70). The tirne course of coronary endothelial dysfunction in acute untreated rejection has been characterized (30) and begins with an endothelial dysfunction due to a progressive destruction of the endothelial lining 5 days after heart transplantation which initially involves Gi-protein mediated relaxations. The dysfunction worsens over time and cornes to affect ail endothelial mechanisrns and vascular srnooth muscle. Contradictory resuits have been reported in the relationship between endothelial dysfunction and vessel morphology. Thus, an irnpaired vasomotor response to papaverine induced increased coronary blood ftow has been described in transplant patients with significant abnormalities of the vessel wall (71) while other studies have flot confirrned the relationship between the impairment of endotheliurn-dependent and independent vasodilatation and degree of intimai hyperpiasia (71-73). Therefore, it appears there is no simple reiationship between coronary artery structure and function in transplanted
patients; some normal appearing vessels do flot dilate whule others with
intimai hyperpiasia show good vasodilatation (74).
The endothelium lies in a strategic anatomical position between the
circulating blood and the vascular srnooth muscle ceil. for this reason, the
graft’s endothelium is the prirnary target of the immunological response to
the aliografi due to the expression of alloantigens on its surface in
conjunction with MHC molecules (direct response), thereby activating host
helper T celis (72). Activated host helper T ceils release interleukin-2 that
leads to the proliferation of other alloreactive celis with the secretion of
cytokines (73). In response, the endotheiiurn expresses celi surface antigens
and adhesion molecules (ICAM,VCAM) and secretes chemoattractant
factors to macrophages which then enter the vessel wall along with
lymphocytes which enhance the migration and proiiferation of vascular
smooth muscle ceils. This leads to an endotheiiaÏ dysfunction which
significantly contributes to the increase of platelet-vessel wall interaction,
vasoconstriction, and proliferation of vascular smooth muscle ceils. Under
these conditions, endothelium-dependent vasodilatation i s reduced and
endothelium-dependent constrictor responses are augmented.
Although the immunoiogical hypothesis for endothelial dysfunction
22
factors may trigger endothelial dysftinction (74,75) after transplantation, including cellular and humoral rejection as described above, hyperlipidemia, cytomegalovirus reactivation, and the toxic effects of immunosuppressants such as cyciosporin A. The process of transplantation itself exposes coronary arteries to cardioplegic solutions (76), cold ischernia (77), preservation solutions (78), and reperfusion after implantation (79-$1) which may cause pathologie activation of the endothelium rendering it dysfunctional. Cold ischernic storage can be deleterious to the preserved organ by causing celiular swelling, extracellular edema, cellular acidosis, depletion ofmetabolic substances, reperfusion injury, calcium overload, and endothelial injury. Injury caused by reperfusion is attributed to ischemic ccli sweiling, the no-reflow phenomenon which is the failure of unifonri blood flow to retum to ail ischemic tissue, reperfusion-induced hemorrhage, and oxygen-derived free radicals (82). The role ofischemia and reperfusion is an early, transient cofactor of endothelial injury after transplantation ($3,84). The activation of the microvascular endothelium leads to the production of oxygen free radicals with subsequent activation of leukocytes and macrophages. Activated celis release oxygen radicals and other mediators such as cytokines, proteases, and eicosanoids. These free radicals can scavenge endothelium-derived nitric oxide (NO) or have direct cytotoxic
effects on the endothelium to induce a coronary endothelial dysfunction. The
chief reactive oxygen species is the superoxide anion. Ail celis generate a
basic level of superoxide anion, but the main producers are the mitochondria
(85). In the setting of acute injury, superoxide anion production is rnediated
chiefly by the enzymes xanthine oxidase, nitric oxide synthase, and by
neutrophiis ($6). In ischemia-reperfusion, the effects of NO can be both
beneficial and detrimental. The vasodilatation induced by NO may be
essential for the perftision of organs during shock (87). The main harmful
effects of NO is the production of peroxynitrite. When NO combines with
the superoxide anion, the resultant peroxynitrite is a much stronger oxidant
than either of its constituents resulting in apoptosis or cytotoxicity. Thus,
postischernic reperfusion injury is a network of interactions mediated by a
large variety of oxidative molecules and aggressive mediators Jeading to an
early endothelial dysfunction.
Endothelial dysfunction afier transplantation is a consequence of
initial injury to the endothelium by mechanisms described above. When the
endothelial ceil is injured, it becornes activated and plays a pivotai role in
the development of atherosclerosis and allograft vascular disease.
Endotheiial activation may lead to the induction ofgenes which are normaliy
24
celi can have critical interactions with smooth muscle celis, macrophages, and platelets, ail of which play a criticai role in the atherosclerotic and coronary artery vascuiopathy process (8$). It has been hypothesized that a transcription factor, nuclear factor-kappa B (89), and other signaling pathways induce the inappropriate and darnaging gene expression by activated endothelial celis. The damaged endothelium results in dysfunctional cellular activities manifested as alterations in endothelial permeability, ioss of antithrombogenic characteristics with microthrombi formation, increased adhesion of leucocytes, and release of vasoactive substances and growth factors ($8). The endothelium norrnally provides a barrier to the passage of plasma proteins and growth factors into the media, but in the presence of endotheliai injury, proteins pass through the endotheliai layer and accumulate in the media (90). Growth factors then stimulate smooth muscle cells to migrate into the intima and proliferate, leading to progressive intimal hyperpiasia (91). Ail of these consequences further promote the developrnent of atheroscierosis and ailografi vascuiopathy.
Preservation Solutions and the Endothelium
Despite the ongoing research in order to extend the length of preservation of organs for transplantation, the ciinicaliy acceptable ischemic
time for heart transplantation remains 4-8 hours (92,93), an amount of time significantly less than for liver (94) and kidney (95) transplantation which are 24 and 48 hours respectively. A retrospective analysis ofheart transplant centers in the United Network of Organ Sharing (96) showed that 167 different preservation solutions are used in heart transplantation today and only 11% of centers use University of Wisconsin solution (UW). Less than optimal solutions may be sufficient if the cold ischernic period is kept < 4
hours but this is rarely the case in a clinical setting. A multivariate analysis of data from the International Heart Transplant Registry (97) showed that ischemic time has a significant (P<O.001) effect on 1 and 5 year mortality. Trnprovements in myocardial preservation solutions can increase the safety period of cold ischemic time without increasing the risk of mortality, and may improve short and long-tenri cardiac function after transplantation. Superior myocardial preservation may also increase the donor pool by allowing the use of suboptimal donors.
Preservation solutions can prevent endothelial ceils from injury from oxygen-derived free radicals during the ischemic period. Niisson et al showed that there was a nonselective decrease in endothelium-dependent relaxations imrnediately afier perfusion with crystalloid cardioplegia with functional recovery after 5 hours of cold storage whereas comparable
26
changes in function of the endothelium were flot observed afier perfusion
with a blood cardioplegic solution (98). The type of solution used for graft
preservation lias been suggested to play a role in the subsequent
development of graft coronary atherosclerosis: specifically, grafts preserved
with an intracellular-type (potassium 125 mEq/L) University of Wisconsin
solution were more susceptible to developing accelerated atherosclerosis
over a 36-month follow-up period than those preserved in extracellular-type
(potassium 30 mEq/L) Stanford solution (99).
As previously rnentioned, there is a four-step sequence of potentially
damaging events on the endothelium in hearts undergoing the transplantation
procedure: (1) donor arrest, (2) cold storage, (3) warm ischemic arrest
associated with implantation, (4) reperfusion (100). Two types of solutions
have traditionally been used to preserve the allograft. The first solution is
used as a flush-out and storage medium during steps 1 through 3 and the
second serves as reperftisate at the onset of step 4. There are limitations
however to the approaches currently used to preserve cardiac allografts. The
first solution used is normalÏy based on the principles of myocardial
preservation that have been developed for cardiac procedures such as
coronary artery bypass surgery. An example is the use of St- Thomas’
necessarily hannful for the allografi but does flot take into account the specificities of the allografi compared to hearts subjected to a single crossclamp period during bypass surgery. These specificities include longer period of ischemic injury, a greater degree of cooling during arrest, and the potential for myocardial injury before harvesting by underlying cerebral injury. The second type of solution utilizes the principles ofpreservation that have been developed for transplantation of abdominal organs. This is illustrated by the use of solutions such as Euro-Collins and University of Wisconsin (UW) in heart transplant recipients. These solutions have yielded good results but their use does flot meet the specificities of an ischernic/reperfused myocardiurn. Differences between the heart and other organs relates to cation transport (101), membrane permeability to different solutes (102,103), and defense mechanisms against oxygen free radicals (104). A solution which is effective for storage may flot be suitable for perfusion during initial arrest (1 05,106) or subsequent cardiopÏegia given
afier reimplantation (107).
These considerations have led Menasché and colleagues (10$) to develop an original preservation solution called Celsior which meets the requirernents for heart transplantation (Table Ï). Celsior combines the principles of organ preservation and allows cold storage specific for
28
ischemic/reperfused myocardium and permits a single solution to be used
throughout ail phases of the transplantation procedure. Celsior is a high
sodium, iow potassium solution which uses glutamate as an energy
substrate. II is an extracellular-type solution with considerable antioxidant
properties due to its high content in reduced glutathione, which confers
superior preservation of myocardial function by preventing production of
free radicals during preservation and reperfusion (105-107, 109). The
prevention of free radical injury is achieved by the combination of three
components: reduced glutathione, mannitol, and histidine. The decrease in
production of active oxygen species, which are known to degrade NO, could
improve protection of endotheliurn-derived NO, which exerts a reguiatory
and protective role on the vascular wall. Celsior also bas the ability to
prevent cellular swelling by the combination of lactobionate (a chelator of
Fe2+ ion) and mannitol, giving it a concentration of 140 mrnol/L to
counterbalance intracellular osmotically active molecules. Experimentally,
hearts arrested and stored in Celsior were found to incur significantly
smalÏer losses of compliance afler ischemia and greater contractile function
during reperfusion compared to hearts exposed to St-Thomas’ Hospital
solution (10$). Another study showed that hearts stored in Celsior for $
minutes had a better lefi ventricular developped pressure than those stored in
St-Thornas’ Hospital, University of Wisconsin (UW), and rnodified (lower
potassium concentration) UW (110). Celsior was also found to significantly
better preserve high energy phosphate levels and myocardial pH versus UW
in canine hearts after 12 hour cold storage with a superior preservation of
glycogen granules at histology (111). Wamecke et al recently published a
report comparing the UW solution to Celsior in a porcine allogeneic heart
transplantation model with measurernent of right ventricular function (RV)
(112). In this study, myocardial preservation with CeÏsior solution resulted
in significantly better post-ischemic RV function than the UW solution afier
transplantation (112). In addition endothelial celis incubated for 6 or 24
hours at 4°C with Celsior or UW solutions were significantly better
preserved in terms of their viability and proliferative capability cornpared to
those incubated in Euro-Collins or St-Thomas solutions (113). Endothelium
dependent relaxations to acetylcholine were also significantly more impaired
in segments of proximal and distal coronary arteries perfused and incubated
for 15 hours in Plegisol solution compared to Celsior (114) in an in vitro
isolated rat heart model. Clinical efficacy studies are currently underway in
Europe and North America (114). Resuits of a prospective multicenter study
30
mortality of 8.6% and a 5-year actuarial survival rate of 75%, comparing
favorably with the international standards set for heart transplantation (114).
Atrial fibrillation and heart block rates were also very low in this cohort,
suggesting a role for Celsior in preserving cardiac sinus rhythrn (114).
As noted above, there is evidence that early development of
endothelial vasomotor dysfunction predicts the subsequent development of
cardiac allografi vasculopathy at 1 year post-transplant (115). The first study
presented in this memoir suggests that storage with the Celsior solution
better preserves endothelial function in cardiac allografts before
reimplantation when compared with a saline solution. The effects of the
Celsior solution on endothelium-dependent and independent relaxations
early (1 rnonth) after cardiac transplantation had not been evaluated.
Therefore, the second study presented in this memoir examines the effects of
Celsior, blood and crystalloid cardioplegic solutions on endothelium
dependent and independent relaxations in porcine coronary arteries 30 days
afier transplantation in a porcine heterotopic retroperitoneal transplantation
model to demonstrate a potential protective role of Celsior on the
endothelium and the development of intimal hyperpiasia at a distance in time
Article #1:
Improved Preservation of Coronary Endothelial
Function with Celsior Compared witli Blood ami
Crystalloid Solutions in ileart Transplantation
E. Dumont, L.P. Perrault, O. Malo, M. Pellerin, M. Carrier Montreal Heart Institute, Montreal, Canada
32
Background: Endothelial injury from preservation solutions has been irnplicated in acute coronary vasospasm and pathologie activation of the endothelium, which can contribute to the development of graft coronary vasculopathy after heart transplantation. Preservation solutions with powerful antioxidant capacity may decrease the occurrence of these complications. This study was designed to evaluate the effects ofCelsior (an antioxidant solution specifically designed for cardiac preservation) in a model of heart preservation (4 hours to mimic clinical conditions) cornpared to two cornmonly used cardioplegic and preservation strategies on coronary artery endothelial function in a porcine model.
Methods: Pig hearts were either imrnediately explanted and preserved (control) or subjected to crystalloid, blood, or Celsior cardioplegia foÏlowed by 4-hour preservation in either saline or Celsior solutions at 4°C (seven experimental groups). Endothelium-dependent relaxation of normal porcine epicardial coronary arteries to serotonin (HT, an agonist that activates 5-HT 1d receptors coupled to Gi proteins) and bradykinin (BK, which activates
B2 receptors coupled to Gq proteins) was studied in standard organ chamber experiments.
Resuits: Cardioplegia with blood, crystalloid, and Celsior solutions caused statistically significant decreases in endothelium-dependent relaxations to
5-HT and BK cornpared to controls. Arteries submitted to cardioplegia and preservati on demonstrated significantly decreased relaxations to 5-HT compared with the control group except the Celsior + Celsior group.
Endothelium-dependent relaxations to bradykinin were unaltered in ail groups afier a 4-hour storage period irrespective of preservation solution used except in one group.
Conclusion : Cardioplegia alone caused a decrease in endothelium dependent relaxations to Gi and Gq-protein coupled agonists. Celsior cardioplegia and storage allowed functional recovery of endothelium dependent relaxations to serotonin cornpared to saline preservation. The observed effect could improve both short- and long-term outcorne in heart transplantation.
34
Introduction
Heart transplantation is the preferTed treatrnent of rnedically unresponsive terminal cardiac failure but rernains associated with a 5-year survival rate of 70% because of the developrnent of graft coronary vasculopathy which leads to suent infarcts, graft failure, and sudden death with no efficacious treatment currently available (1-3). Coronary endothelial dysfunction precedes the development of atherosclerosis in many vascular conditions such as diabetes mellitus, hypercholesterolemia and is predictive in hurnan transplant recipients of the developrnent of intimal hyperplasia and of adverse outcornes (4). Nurnerous factors may trigger the endothelial dysfunction (5,6) afier transplantation including cellular and humoral rejection, hyperlipidemia, cytomegalovirus reactivation, the toxic effects of cyclosporin. The process of transplantation itself with exposure of coronary arteries to cardioplegic solutions (7), cold ischernia (8), preservation solutions (9), and reperfusion afier implantation (10-12) may cause pathological activation (13) of the endothelium rendering it dysfunctional. Endothelial dysfunction is associated with an attendant loss of the endothelial celis regulatory and protective properties on the homeostasis of the vascular wall which is achieved, under normal conditions, by the release of endothelium-derived relaxing factors (EDRF) such as nitric oxide (NO),
prostacyclin (PGI2) and the endothelium-derived hyperpolarizing factor
(EDHF). Under pathological conditions, the endothelium may also release
endotheliurn-derived constricting factors (EDCF) such as the superoxide
anion (Q2.), endoperoxydes, thromboxane A2 and endothelin- 1 which may contribute to the developrnent of atherosclerosis by inactivating EDRFs or
by a direct effect on the vascular wall. Ischernia and reperfusion lead to the
production of ftee radicals which can scavenge endotheÏiurn-derived NO and
induce a coronary endothelial dysfunction rnanifested by reduced dilatory
responses to endothelium-dependent agonists (14).
Preservation solutions could prevent endothelial ceils from injury due to
oxygen derived free radicals during the ischernic period. Nilison et al.
showed that preservation for 5 hours in a crystalloid solution caused a
reversible acute endothelial dysfunction while no alterations were detected
when hypothermic blood cardioplegia was used (15). The type of solution
used for graft preservation has been incriminated in the causation of a higher
incidence of graft coronary atherosclerosis: specifically, grafis preserved in
the intracellular type University of Wisconsin solution were more
susceptible to developing accelerated atherosclerosis than those preserved in
the extracellular type Stanford solution (16). Moreover, a link has been
36
duration of cold ischernia and the subsequent developrnent of graft coronary vasculopathy (17,1 8).
There is substantial experimental evidence which shows that oxygen derived free radicals contribute to earÏy postoperative failure of cardiac allografts. The Celsior solution (Pasteur-Mérieux, Lyon, France) is an original preservation solution specifically designed for prolonged heart preservation (Table 1). It combines the general principles of preservation solutions as well as aspects specific to the heart such as prevention of the development of ischernic contracture and dysfunction due to ederna. It is an extracellular type solution with a great antioxidant capacity due to its high content in reduced glutathion which confers a superior preservation of myocardial function by preventing production of free radicals during preservation and at the time of reperfusion (15,22-24). The Celsior formulation was developed with two main principles in mmd: first, the prevention of cellular swelling which is achieved by the combination of lactobionate (a chelator of Fe ion, implicated in the Findel reaction) and mannitol (which has a smaller antioxidant effect by scavenging 0 singlet) which has a concentration of 140 mrnol/L to counterbalance intracellular osmotically active molecules. The second guiding principle is the protection from free radical injury which is achieved by the combination of three
components: reduced glutathion, mannitol and histidine. These effects on
reactive oxygen species production, which are known to degrade NO, could
improve preservation of endotheÏial-derived NO. Celsior was found to be
effective in vitro in an isolated rat heart model and in vivo in a rabbit
heterotopic heart transplantation model (25). Clinical efficacy studies are
currently underway in Europe and North Arnerica, but the effects of Celsior
on endothelial function and on the developrnent of grafi vasculopathy are
unknown.
This study was designed to investigate the effect of cardioplegia and
prolonged hypothermie storage with Celsior, blood or a crystalloid solution
(KC1 60 mEq added in) followed by 4 hours ofpreservation at 40
C in saline (NaC1 0.9%) or Celsior solutions, on the endothelial function of porcine
epicardial coronary arteries.
Materials
ami
MetliodsA. Animais and Experirnental Groups
Forty Landrace piglets with a mean weight of 20.2 ± 2.0 kg (range
14.0 to 22.2 kg), age 10 + 1 weeks, were used for investigating the effects of
3$
function. Ten groups were studied (Table 2): hearts in the control group
were submitted to immediate excision without administration ofcardioplegia
nor preservation (Group 1); three groups had cardioplegia with either Celsior
(Group 2), blood (Group 3), or crystalloid (Group 4) solutions without
preservation; two groups had cardioplegia induced with a crystalloid
solution and were then stored for 4 hours in a saline solution (0.9% NaC1)
(Group 5) or 4 hours in Celsior solution (Group 6) (4°C); two groups had
cardioplegia induced with normothermic blood cardioplegia (37 C) and were
then stored for 4 hours in saline solution (Group 7), or 4 hours in Celsior
solution (Group 8); finally two groups underwent cardiopiegia with Ceisior
and were stored in either saline (Group 9) or Celsior solutions (Group 10)
for 4 hours. Ail animais received humane care in cornpiiance with the
recommendations ofthe guidelines on the care and use oflaboratory animais
issued by the Canadian Councii on Animais and Guidelines ofAnimai Caret,
and were approved by a local committee.
B. Anesthesia and Surgical Technique
Anesthesia was induced with an intramuscuiar mixture injection of
ketamine (20 mg/kg; Rogarsetic, Montreal, Quebec, Canada) and xylazine
ventilated with an 02/air mixture (3:2). Respiratory control was maintained by frequent determinations of arterial blood gases and acidosis was balanced with 8.4% sodium bicarbonate (Abbott laboratories, Montreal, Que., Canada). Light anaesthesia was supported by halothane 1% (halocarbon laboratories, Markham, NJ, USA). Hair was shaved at the operative field and the skin was surgically disinfected. A median stemotorny and pericardial incision was performed, and heparin (3 mg/kg, heparin sodium; Leo laboratories, Kingston, Ont., Canada) was given through a central venous catheter. The distal right innominate arteiy was ligated and the proximal portion cannulated with a polyvinyl chloride catheter positioned in the ascending aorta for administration ofcardioplegia. Afler clamping ofthe aortic arch between the two innorninate arteries, asystole was induced through injection of the cardioplegia solution in the ascending aorta at a maximal pressure of 60 mm Hg. The heart was vented by incision of the right and left atrial appendages and the heart was excised.
C. Cardioplegia and Storage Period
In the control group, the hearts were irnmediately excised and transferred to organ chambers. The cardioplegic solutions used were crystalloid (12 mL/kg), Celsior (12 mL/kg) or normothermic blood (4:1 blood-Ringer’s
40
lactate ratio 15 mL/kg). In the normothermic blood group, blood was harvested during the initial phase of anaesthesia in a 450 mL bag with 63 mL citrate dextrose phosphate and kept normothermic. Potassium chloride (mean 60 mEq) was then added through a Y-tubing in the crystalloid, blood and Celsior solution groups and administered in the aortic root. Afier ftooding ofthe pericardium with cold storage solution, the heart was excised and placed in the designated storage solution for 4 hours at 40
C or transferred to organ chambers as described above.
D. Vascular Reactivity
The lefi anterior descending, left circumilex and the right coronary arteries were excised from the myocardium and dissected free from the epicardium, rnyocardium and from adventitial tissue and divided in 4 mm wide rings and used randornly in organ chamber experirnents. Sixteen rings were used for each animal, eight rings per endothelium-dependent agonists (5-HT and BK).
The endothelial function of control and arteries rings submitted to cardioplegia and storage was studied in organ chambers fflled with Krebs bicarbonate solution (20 mL at 37° C: composition in mmol/L: NaCY 118.3, KCÏ 4.7, MgSO4 1.2, KH7PO4 1.2, glucose 11.1, CaCl2 2.5, NaHCO3 25 and
calcium ethylenediaminotetraacetic acid 0.026). Oxygenation was insured using a 95% 02 / 5% C02 gas mixture. The rings were suspend between two metal stirrups, one of which was connected to an isometric force transducer. Data were collected with a data acquisition software (10S3, Emka Inc., Paris, France).
Ail studies were performed in the presence of indomethacin (1 0 mol/L, to exciude the production of endogenous prostanoids), propranolol (1 0 rnol/L, to prevent the activation of f3-adrenergic receptors), and ketanserin (incubated 45 minutes before the addition ofserotonin: 106 mol/L, to block serotonin 5HT-2 receptors which cause contraction of smooth muscle ceils in the absence of ketanserin). Each preparation was stretched to the point of its active length curve (usually 3.5 g), as determined by measuring the contraction to potassium chloride (30 mrnol/L) at different levels of stretch, and then stabilised for 30 minutes. The maximal contraction was determined with potassium chloride (60 mmol/L) and rings were excluded if they failed to contract to potassium chloride (exclusion rate less than 5%). After washing and 45 minutes of stabilisation, prostaglandin F2Π(range, 2 X
1 OE6 to 1 0 mol/L) was added to achieve a contraction averaging 50% of the maximal contraction to KCÏ (60 mmol/L). Endothelium-dependent relaxations to serotonin (5-HT, i0’° to i0 mol/L, an agonist which
42
activates 5HTJd receptors coupled to Gi-proteins which lead to the release of NO) and bradykinin (BK, 1O2 to 106 mol/L which activates B2 receptor coupled to Gq-proteins which lead to the release of NO and EDHF) were rneasured. At the end of the experirnents, endothelium-independent relaxations were studied using a single bolus of sodium nitroprusside (SNP iO rnol/L, a nitric oxide donor). No rings were exposed to more than one endothelium-dependent agonist.
E. Drugs
Ail solutions were prepared daily. Bradykinin, 5-Hydroxytryptamine creatinin sulfate (serotonin), indomethacin, ketanserin and sodium nitroprusside were purchased from Sigma Chemical Co. (Oakville, Ontario, Canada). Propranolol was purchased from Biomol Research Laboratories Inc. (Plyrnouth Meeting, PA, USA) and prostaglandin F2Πwas purchased from Cayman Chemical Company (Ann Arbor, MI, USA).
F. Statistical analysis
Contractions to PGF2Πare expressed as the percentage ofthe maximal contraction to potassium chloride (60 mmol/L) for each group and expressed
as means ± standard error of the mean (SEM); an” refers to the number of
animais studied. Relaxations are expressed as percentage of the maximal contraction to PGF2Πfor each ring. ANOVA studies were performed to compare dose-response curves. Differences were considered to be statistically significant when the value ofp < 0.05.
Resu its
A. Vascular Reactivity Contractions
There were statistically significant differences in the amplitude of the contraction to potassium chloride (60 mmol/L) between the crystalloid/Celsior cardioplegia groups and the control/blood groups. There were also statistically significant differences in the amplitude of the contraction to potassium chloride (60 rnmol/L) among the groups, except in arteries subrnitted to cardioplegic arrest with Celsior and preservation in saline (group 6) which showed a significant decrease (f igure 1). There were no statistically significant difference in contractions to PGF2Π(range 2 i o6 to 1 0 mol/L
)
except in the group subrnitted to cardioplegic arrest with Celsior and preservation in the saline solution (group 6) which showed a significant increase ($8.1 ± 4.3%,p < 0.05 vs all groups) (Figure 2).44
Relaxations
$erotonin. There were no significant difference in endothelium-dependent relaxation to serotonin and bradykinin between arteries submitted to storage in saline only (group 1) versus normal porcine coronary arteries freshly explanted and tested (data not shown).
There was a statistically significant decrease in endothelium-dependent relaxations to serotonin in porcine coronary arteries submitted to crystalloid, blood, or Celsior cardioplegia alone (Groups 2-4) (Figure 3).
There were no statistically significant decreases of endothelium-dependent relaxations to serotonin in the Celsior cardioplegia + Celsior preservation
group (group 10) compared with controls. When the Celsior and crystalloid solutions were used for cardioplegia , the arteries subsequently preserved in
Celsior (group 6 and 10) showed a significantly superior preservation of relaxation to 5-HT compared to those preserved in saline (p <0.05 for Celsior
+ Celsior group vs Celsior+ Saline only) (Figure 4). There was a statisticaliy
significant decrease in endothelium-dependent relaxations to serotonin in arteries from ail other cardioplegic/preservation groups (groups 5,6,7,8,9) compared with controls (Figure 4).
Bradykinin. There was a statistically significant decrease of endothelium dependent relaxations to bradykinin in rings from ail coronary arteries exposed
to either blood, crystalloid, or Celsior cardioplegia alone (figure 3). There was a statistically significant decrease of endothelium-dependent relaxations to bradykinin in rings from the Blood + Celsior and Celsior + Saline groups
(groups 8 and 9) at one concentration
(
3 x 10 rnol/L) (p <0.05 vs control group). There were no statistically significant differences between ail other four cardioplegic/storage groups compared with controls (figure 5).Endothetium-independent reÏaxations. There were no statistically significant differences in the maximal relaxation to sodium nitroprusside (SNP;
i0 mol/L) in ail groups (data not shown).
Discussion
The major findings of this study are that exposure of porcine coronary arteries to cardioplegic arrest with crystalloid, blood or Celsior cardioplegic solutions cause a significant decrease in endothelium-dependent relaxations to serotonin and bradykinin. Subsequent preservation in saline does flot provide recovery of endothelium-dependent relaxations. Induction of arrest and preservation with Celsior allows for functional recovery of endothelium dependent relaxations to both serotonin and bradykinin and thus beller preserves the endothelial function of coronary arteries under the present experimental conditions.
46
Endothelium-dependent contractions
The significant decrease in the amplitude of contraction to potassium chloride
in the arteries submitted to cardioplegia with Celsior and saline preseiwation
may be related to the greater potassium loading or decreased sensitivity to
exogenous potassium in this group although it was not observed in the other
groups in which Celsior was used for cardiopiegia. The greater amplitude of
contraction to prostaglandin F2Πin the arteries subrnitted to Celsior
cardioplegia and saline preservation may reflect a greater sensitivity of the
vascular srnooth muscle ceil to exogenous prostaglandins. Under certain
conditions, endothelium-dependent contractions can be explained either by
the withdrawai of the release of endotheliurn-derived relaxing factors (NO,
prostacyciin) or by the production of diffusibie vasoconstrictor substances
coined endotheiiurn-derived contracting factors (EDCF).
Effect of the cardioplegic solution
Perfusion of coronary arteries with the crystalloid, blood, and Celsior
cardiopiegic solutions ail caused a decrease in endothelium-dependent
relaxations to serotonin and bradykinin. Perfusion of coronary arteries with
the crystalloid and blood solutions followed by preseiwation in either saline or
Celsior to obtain an electrornechanical silence induces a decrease in Gi
followed by preservation in saline also causes a sirnilar type of dysfunction. This endothelial dysfunction preferentially involves Gi- protein mediated relaxations as demonstrated by reduced responses to serotonin which acts by binding to 5-HTID receptors coupled to Gi-proteins in normal porcine coronary arteries. The Gq-protein rnediated relaxations are altered to a lesser degree as evidenced by preservation of relaxations to bradykinin (which binds to B2 receptors coupled to Gq-proteins and leads to the release of NO and EDHF in normal porcine coronary arteries) except at one concentration in the group submitted to crystalloid cardioplegia with storage in the saline solution. The G• protein mediated pathway is one of the first pathways affected in many cardiovascular diseases such as acute and chronic cardiac rejection (19,26), atherosclerosis (27,28) and may be an early marker ofpathological endothelial activation implicated in the developrnent ofvascular diseases. Preservation of this sensitive endothelial celi signaling pathway, reflecting the final capacity of the vascular wall to produce the vascular protector endothelial NO is important to avoid progression to a generalized endothelial dysfunction associated with advanced atherosclerosis. Endothelium-dependent relaxations rnediated by the NO donor, sodium nitropmsside, were unaffected, therefore demonstrating the integrity of vascular smooth muscle cells. This confirms that the reduction of
48
vascular relaxation observed throughout the groups studied is due to endothelium-dependent mechanisrns.
Perfusion of coronaries with a crystalloid solution is known to alter endothelium-dependent relaxations in comparison to perfused controï (29) without any morphological alterations of endothelial cells. Cardioplegia with St-Thomas solution delivered at 60 to 65 mm Hg significantly reduces endothelium-dependent relaxations in addition to the cold ischemic period and reperfusion (30). However, others have shown that use of a hyperkalernic crystalÏoid cardioplegic solution does not irreversibly alter endothelium dependent relaxations to acetylcholine (G-protein rnediated) or adenosine diphosphate (ADP) (31). Perfusion at a pressure of 240 cm H20 impairs the vasodilatory response to serotonin and results in vasoconstriction (32). These studies suggest that the dysfunction observed are due to a traurnatic effect of baroreceptors or to shear stress on the vascular wall ofthe arteryand not from the cardioplegia solution itself. He et al. demonstrated that the relaxation to EDHF is altered following exposure to a hyperkalemic solution due to partial membrane depolarisation of the srnooth muscle cells affecting the potassium channels and the release of the endotheliurn-derived relaxing factors (34,35).
Although Celsior is sufficient in itself to induce cardioplegic arrest, potassium was added in the Y-tubing as with the two other cardioplegic
solutions to negate the confounding effect of hyperkaliemia on the induction ofcoronary endothelial dysfunction in this mode!.
Effect of the storage so]ution
The preservation times currently considered safe range between 4 to 6 hours and longer ischernic periods are associated with a reduction ofthe survival rate of heart transplant recipients (35). Storage at 40
C was superior in preserving the endothelial and myocardial function cornpared to 200 C in an iso!ated heart
model in rats (36). Accordingly, porcine coronary arteries exposed to 4 hours of storage at 40
C dernonstrated similar endothelium-dependent relaxations compared with fresh!y harvested arteries. This strongly suggests that the endothelial dysfunction observed in the groups exposed to cardioplegia solutions is not a consequence of preservation but a resu!t of perfusion of the coronary arteries during cardiac arrest (37). The preservation time of 4 hours in this study was chosen since it represents the average ischemic tirne in a clinica! heart transplantation setting.
Storage of the heart in a saline solution does flot alÏow recoveiy of Gi protein mediated relaxations during the 4 hour storage period as shown by the decrease relaxations to serotonin in the cardioplegia groups stored in saline. Storage in the Celsior solution was associated with a better preservation ofGi protein mediated relaxations except in the group in which blood was used as
50
the cardioplegic solution which showed a significant endothelial dysfunction.
The fact that storage in saline after Celsior cardioplegia allowed for recovery
of endothelial function to Gq-protein mediated relaxations can be explained in
a twofold maimer. Firstly, it is highly likely that endothelial dysfunction seen in ail three groups afler cardioplegia alone was due to shear forces and
barotrauma and not to the cardioplegic solution itseif It is known that infusion
of a hyperkalemic crystalloid cardioplegic solution does flot alter the
endothelium in a irreversible manner (31) and thus the 4 hours of preservation
in saline could have allowed for functional recovery of the endothelium in the
more resistant Gq-protein rnediated pathway. Secondly, the Celsior
cardioplegic solution rnight cause initial dysfunction but may aÏlow a better
recovery ofendothelial function regardless ofthe preseiwation solution used by
its beneficial properties likely related to its antioxydant capacity to eliminate free radicals production during ischemia combined to a reduction of
myocardial oedema (25). Use of this new preservation solution could protect
grafis from the deleterious effects of preservation (38) as suggested by a
clinical study in which use of Celsior was associated with an increased
inotropic and vasodilatory response in transplanted patients in contrast to the
University of Wisconsin solution (39). This is consistent with the observation
conditions is use of Celsior for induction of cardioplegia and storage in Celsior. Indeed, endothelium-dependent relaxations of coronary arteries exposed to cardioplegia and preservation with Celsior solution are sirnilar to the control group. Storage of the heart for 24 hours in a crystalloid solution with the antioxydant indonoindole also preserves endothelium-dependent relaxations in comparison to crystalloid alone (40) suggesting that improvernent in the formulation of preservation solutions may lengthen the safe ischemic tirne for heart transplantation. Lung preservation, using Celsior, (41,42) has also yielded favorable initial results.
Limitations
Damage by production of free radicals by leukocytes at the tirne of reperfusion is a significant contributing factor in the developrnent of pathological activation of the endothelium (14). The present study solely addresses the effects of arrest and preservation solutions on coronary endothelial function. Further studies evaluating the consequences of myocardial reperfusion following storage on coronary endothelial function are needed since the endothelial dysfunction observed with arrest and storage are likely to be compounded by reperfusion. The benefits of using Celsior may become even more evident considering its proven capacity for capture of oxygen derived ftee radicals occurring during reperfusion. Use of
52
normal swine does flot reftect the clinical condition of organ harvest associated with brain death which may be associated with massive cathecolamines release and functional disturbances depending on the rnechanism of death. Experimental studies evaluating the effects of this condition on rnyocardial and coronary endothelial function may yield further insight on the problematic of solid organ procurernent and storage. Finally, studies of the effects of preservation solutions and storage on coronary microvascular function should also be undertaken considering the central role ofthe microcirculation in autoregulation of coronary blood flow.
Conclusion
In summary, cardioplegic arrest and storage with Celsior constitutes the best strategy permitting the preservation of endothelium-dependent relaxation in porcine epicardial coronary arteries. $uperior preservation of coronary endothelial function at the time of implantation may reduce early grafi failure by improving coronary blood flow and limit the development of graft coronary vasculopathy by maintaining the protective effects of the endothelium on the vascular wall.
Bibliography
1. Davis S, Yeung A, Meredith I, Charbonneau F, Seiwyn P, Anderson T. Early endothelial dysfunction predicts the developrnent of
transplant coronary artery disease at 1 year posttransplant. Circulation 1 996;93 :45 7-62.
2. Billingharn M. Cardiac transplant atherosclerosis. Transplant Froc 1987;19(suppl 5):19-25.
3. Young J. Allografi vasculopathy: diagnosing the nernesis of heart transplantation. Circulation 1 999;3 :458-60.
4. Libby P, Salornon R, Payne D, Schoen F, Pober J. Functions of vascular wa!l celis related to development of transplantation associated coronary arteriosclerosis. Transplant Froc 1989 ;2 1:3677-84.
5. Drexier H, FischelÏ TA, Pinto FJ, Chenzbraun A, J B, Cooke JP, Alderman EL. Effect of L-arginine on coronary endothelial function in cardiac transplant recipients, relation to vesse! wa!l morphology. Circulation. 1994:1615-1623.
6. Ross R. The pathogenesis ofatherosclerosis an update. NEngi JMed. 1986;3 14:488-500.
54
7. Lticher IF, Vanhoutte PM. The endothelium: Modulator of cardiovascular function. In: Press C, ed. Boca Raton, FL; 1990:1-228. 8. Furchgott R, Vanhoutte P. Endothelium-derived relaxing and
contracting factors. FA$EB 1 1 9$9;3 :2007-201$.
9. Dorphin A. Nucleotide binding proteins in signal transduction and disease. Trends Neurosc. 19$7;10:53-57.
10. Rickenbacher P, Pinto F, Lewis N, Hunt S, Alderman E, Schoroeder F, $tinson E, Brown B, Valantine H. Prognostic importance of intimai thickness as rneasured by intracoronary uitrasound after cardiac transplantation. Circulation. 1995;92:3445-3452.
11. Mugge A, Brandes R, Heublien B, Nolte C, Haverich A, lichtlen P. Endothelial dysfunction in heart transplanted patients with graft vasculopathy. EurHeartJ. 1995;16 (Suppl J):78-$3.
12. Flavahan NA, $himokawa H, Vanhoutte PM. Pertussis Toxin Inhibits Endothelium-dependent RElaxations to Certain Agonists in Porcine Coronary Arteries. Journal ofPhysioÏogy. 1 9$9;40$ :549-560.
13. Flavahan N, Vanhoutte P. Endothelial ceil signaÏing and endotheliai dysfunction. Am J Hypertension. 1995;$ :28S-415.
14. $ellke FW, Shafique T, Ely DL, Wientraub RM. Coronary Endothelial Injury After Cardiopulmonary Bypass and Ischemic
Cardioplegia Is Mediated By Oxygen-Derived Free Radicals. Circulation. 1993;8$ part 2:395-400.
15. Nilsson FN, Miller VM, Vanhoutte PM, McGregor CGA. Methods of cardia preservation alter the frmnction of the endothelium in porcine coronaiy arteries. The Joïtrnal od Thoracic and Cardiovascular Surge;y. 199 1;102 No 6:923-30.
16. Drinkwater DC, Rudis E, Laks H, Marino J, Stem D, Ardehali A, Aharon A, Moriguchi J. University of Wisconsin Solution versus Stanford Cardioplegic Solution and the Development of Cardiac Allograft Vasculopathy. JHeartLung Transplant. 1995;14 No 5:89 1-6.
17. Gaudin P, Rybum K, Hutchins. Peritransplant injury to the myocardiurn associated with the developrnent of accelerated arteriosclerosis in heart transplant recipients. Am J Sïtrg Pathol. 1994;18:338-346.
18. Veinot J, Walley V. Cardiac transplant arteriopathy. Am J Surg Pathol. 1995;19:727-734 (Letter).
19. Perrault L, Bidouard J, Janiak P, Villeneuve N, Bruneval P, Vilaine J, Vanhoutte P. Impairment of G-protein-mediated signal transduction in
